Slurp pipeline

- Delete: Initialized a variables for 'all relationships' and 'term parent map', but did a more OAK-reliant refactor and removed these.
- Add: Term class
- Update: Big refactor to utilize Term class
- Add: Slurp output TSV files
- Add: utils.py, which includes Term class
- Add: New param: max_id

Misc
- Update: run.sh: This change will allow any Python package updates to be retained.
- Add: SPARQL jinja query to get parents. Used by slurp pipeline.

src/ontology/mondo-ingest.Makefile

@@ -344,16 +344,21 @@ slurp/:
 	mkdir -p $@
 
 # min-id: the next available Mondo ID
+# TODO: `pip install` stuff is temporary until ODK docker up to date w/ recent OAK updates
+# TODO: Check if removing --rf from run.sh will fix need for pip install. havent been able to try yet; docker issue
 slurp/%.tsv: $(COMPONENTSDIR)/%.owl $(TMPDIR)/mondo.sssom.tsv $(REPORTDIR)/mirror_signature-mondo.tsv | slurp/
+	# pip install --upgrade -r $(RELEASEDIR)/requirements-unlocked.txt
 	python $(SCRIPTSDIR)/migrate.py \
 	--ontology-path $(COMPONENTSDIR)/$*.owl \
 	--sssom-map-path $(TMPDIR)/mondo.sssom.tsv \
 	--onto-config-path metadata/$*.yml \
-	--min-id 123000 \
+	--min-id 850000 \
+	--max-id 999999 \
 	--mondo-terms-path $(REPORTDIR)/mirror_signature-mondo.tsv \
+	--slurp-dir-path slurp/ \
 	--outpath $@
 
 slurp-%:
-	$(MAKE) slurp/%.tsv
+	$(MAKE) slurp/$*.tsv
 
-slurp-all: slurp-omim slurp-doid slurp-ncit slurp-ordo slurp-icd10cm slurp-icd10who
+slurp-all: slurp-omim slurp-doid slurp-ordo slurp-icd10cm slurp-icd10who slurp-ncit

src/ontology/run.sh

@@ -45,7 +45,7 @@ if [ -n "$USE_SINGULARITY" ]; then
         -W $WORK_DIR \
         docker://obolibrary/$ODK_IMAGE:$ODK_TAG $TIMECMD "$@"
 else
-    docker run -v $VOLUME_BIND -w $WORK_DIR -e ROBOT_JAVA_ARGS="$ODK_JAVA_OPTS" -e JAVA_OPTS="$ODK_JAVA_OPTS" --rm -ti obolibrary/$ODK_IMAGE:$ODK_TAG $TIMECMD "$@"
+    docker run -v $VOLUME_BIND -w $WORK_DIR -e ROBOT_JAVA_ARGS="$ODK_JAVA_OPTS" -e JAVA_OPTS="$ODK_JAVA_OPTS" -ti obolibrary/$ODK_IMAGE:$ODK_TAG $TIMECMD "$@"
 fi
 
 case "$@" in

src/ontology/slurp/icd10cm.tsv

@@ -0,0 +1,25 @@
+mondo_id	xref	label	definition	parents
+ID	A oboInOwl:hasDbXref	LABEL	A IAO:0000115	SC %
+MONDO:850001	ICD10CM:A00-B99			
+MONDO:850002	ICD10CM:C00-D49			
+MONDO:850003	ICD10CM:D50-D89			
+MONDO:850004	ICD10CM:E00-E89			
+MONDO:850005	ICD10CM:F01-F99			
+MONDO:850006	ICD10CM:G00-G99			
+MONDO:850007	ICD10CM:H00-H59			
+MONDO:850008	ICD10CM:H60-H95			
+MONDO:850009	ICD10CM:I00-I99			
+MONDO:850010	ICD10CM:ICD-10-CM			
+MONDO:850011	ICD10CM:J00-J99			
+MONDO:850012	ICD10CM:K00-K95			
+MONDO:850013	ICD10CM:L00-L99			
+MONDO:850014	ICD10CM:M00-M99			
+MONDO:850015	ICD10CM:N00-N99			
+MONDO:850016	ICD10CM:O00-O9A			
+MONDO:850017	ICD10CM:P00-P96			
+MONDO:850018	ICD10CM:Q00-Q99			
+MONDO:850019	ICD10CM:R00-R99			
+MONDO:850020	ICD10CM:S00-T88			
+MONDO:850021	ICD10CM:U00-U85			
+MONDO:850022	ICD10CM:V00-Y99			
+MONDO:850023	ICD10CM:Z00-Z99